Automatic call distributors (ACDs) and other contact or call centers typically include specialized systems designed to match incoming requests for service, for example a telephone call or an e-mail, with a resource that is able to provide that service, for example a human call center agent. ACDs generally perform one or more of the following functions: (i) recognize and answer incoming calls; (ii) review database(s) for instructions on what to do with a particular call; (iii) using these instructions, identify an appropriate agent and queue the call, often times providing a prerecorded message; and (iv) connect the call to an agent as soon as the agent is available.
Hosted ACD call centers enable customers to focus on their core business while reducing capital expenditure and freeing them from the hassle of managing call center resources (e.g., agents and equipment). Such hosted centers may provide equipment, agents and supervisors to handle incoming calls on behalf of customers (clients). As part of the contract between the hosted ACD service provider and their clients, a service level agreement (SLA) may be negotiated. One key aspect of the SLA is meeting predetermined service level goals (SLGs) or targets such as answering a specified percent of incoming calls are within a predetermined time (e.g., 80% of incoming calls must be answered within 20 seconds). The hosted ACD service provider may be paid in accordance with its ability to meet the agreed upon SLA.
Hosted ACD software typically handles incoming calls in a linear first in first out (FIFO) algorithm. More sophisticated systems may provide priority queues for more important customers; but again within each queue, callers are serviced in accordance with the linear FIFO algorithm.
Some systems, such as those from Avaya, aim to improve the service level (SL) by dynamically moving agents between queues. To achieve this functionality, these systems continuously calculate the SL for each queue. If they find that a certain queue is able to meet its target SLA while another queue is falling behind, the system automatically moves agents from the queue that meets the SL to the queue that is falling behind.